ADOLESCENCE

Growing up female: Navigating body image, eating, and depression

Julia A. Graber and Jeanne Brooks-Gunn

Based on part of the Adolescent Study Program at Columbia University, this article contains a description of the dramatic increase in eating and depressive disorders experienced by significant numbers of girls as they pass through puberty. This study was unique in that it followed girls with eating and depressive problems across the entire adolescent decade. The authors discuss characteristics leading to early problems that are predictive of more intense problems over the years. They also suggest interventions that might improve life experiences of adolescents and offer positive lifelong consequences.

Adolescents face many challenges as they make the transition from being children to becoming adults. They must cope with changes at school, new social demands such as dating, and new roles in their families as they assert their independence and increased decision-making skills. Though it has long been thought that adolescents’ moodiness and irritability mean they are plagued by problems, most adolescents experience healthy development and have a minimum of difficulty. The challenge to researchers, parents, and practitioners is to identify those individuals who will have more problematic transitions from childhood to adulthood.

The problems experienced by boys and girls at this time in their lives take different forms. Girls are at much greater risk for developing two types of adjustment problems or disorders: depression and eating problems. Related to these are difficulties with body image. By contrast, boys more often exhibit externalizing or acting-out problems such as conduct disorders and juvenile delinquency

To understand the development and consequences of having a serious depressive, eating, or body-image problem in adolescence, our group conducted studies on girls’ development under the aegis of the Adolescent Study Program (see Note 1). The program has encompassed a series of investigations examining the physical, psychological, and social development of adolescent girls and women.

Since the start of the program nearly 15 years ago, we have worked with over 1,000 girls from ages 10 through 23, their mothers, and their schools, focusing on the challenges of adolescence. In this article, we briefly discuss the rates and nature of eating, depressive, and body-image problems for adolescent girls. We then present findings from the Late Adolescence Study on the development of these problems in a sample of girls seen over an 8-year period, from early adolescence into young adulthood.

In our work, we have focused on understanding the range of experiences from healthy, normal development to unhealthy development in adolescent girls. In particular, we have focused on eating, depressive, and body-image problems, as these are the problems that most often plague them. We distinguish problems from clinical disorders in that girls who experience an eating or depressive problem may not meet strict criteria for a diagnosable disorder but are still experiencing a significant number of unhealthy symptoms and impairment in their lives. We have also found, as we chart girls’ development across the adolescent decade, that girls with early problems often develop full-blown disorders; these are the girls who might benefit most from early detection and intervention.

The transition to adolescence and the emergence of problem behavior
The transition into adolescence has been defined by physical changes of puberty, school changes from an elementary to a middle or junior high school environment, cognitive changes with increased ability to understand cause and effect and think about the future, and changes in family relationships as adolescents seek more independence from parental supervision. Certainly, many of these changes provide opportunities for new and exciting experiences and achievements as adolescents practice adult roles and develop a sense of their independent identity. The biological, social, and personal challenges, however, may be overwhelming for some.

Rates of eating and depressive disorders increase dramatically around the time that girls pass through puberty or in the years just after pubertal development. Estimates of the occurrence of eating disorders, specifically Anorexia Nervosa and Bulimia Nervosa, range from 0.2% for Anorexia to 2.8% for Bulimia (rates used in the Diagnostic and Statistical Manual of Mental Disorders, 4th ed., American Psychiatric Association, 1994). It has generally been accepted, however, that as many as 20% of adolescent girls and young women experience serious eating problems that do not meet all diagnostic criteria. Whereas eating problems are most common in White, middle to upper-middle class girls, rates appear to be increasing in middle class girls of different racial and ethnic backgrounds (see Note 2).

A range of factors that are associated with eating problems and disorders has been identified. Psychological characteristics such as perfectionistic tendencies and family factors such as low warmth or over-controlling behaviors by parents may place girls at risk for developing an eating problem or disorder. In addition, the dramatic changes in physical appearance resulting from pubertal changes such as growth spurt in both height and weight and development of breasts and pubic hair require girls to reevaluate their images of themselves; increases in body fat have been particularly associated with increased dieting. It has been suggested that cultural ideals for thinness that exist in White middle and upper-middle class communities are in direct conflict with the normal increases in body fat that occur during puberty. Not only do girls who are not overweight frequently diet at this time, but we have found that their dieting practices are often unhealthy:

These girls rely on fad diets or ignore basic nutritional requirements. Such behaviors are particularly alarming given that girls’ bodies will continue to grow even after puberty and into adulthood. For instance, bone density increases, resulting in stronger bones and fewer injuries. Most girls with serious eating problems experience disturbances in functioning in several areas of their lives; that is, if they have poor family relationships and perfectionistic tendencies, they may be most at risk for allowing dieting practices to become pathological.

Depressive problems are even more common than eating problems among adolescent girls. Rates for having a Major Depressive Episode range from 7% to 10% of all adolescents, with girls being about twice as likely as boys to have a depressive disorder. Less severe depressive problems are much more pervasive, with as many as 35% of adolescents experiencing a period of seriously depressed mood and increased depressive symptoms (see Note 3).

As with eating problems, depressive problems have been associated with dysfunction in multiple domains of girls’ lives. Again, it has been suggested that pubertal development may be linked to increased moodiness and possibly subsequent depression. In addition, the early years of adolescence also involve many changes in other aspects of girls’ lives. In addition to the change from elementary to secondary school, the increased social demands faced by adolescents in both same-gender and mixed-gender social activities may also be stressful. In fact, our group has found that, for girls, the number of stressful life events peaks around age 14, indicating that young adolescents actually experience more difficult life changes or events than younger children or older adolescents. We found that the experience of more life events was predictive of increases in depressive affect.

Most individuals who interact with girls in early adolescence are not surprised to learn that many of them experience negative feelings about their bodies at this time. Fortunately, poor body image does not persist throughout adolescence for most girls. Although body-image problems are not a distinct psychological disorder, poor body image has been associated with both eating and depressive problems. In fact, a major disturbance in body image is one criterion for an eating disorder diagnosis. Research by Simmons and Blyth (1987) and others has shown that girls have poor body image around the time of puberty, but, fortunately, for most girls, body image actually improves across the adolescent decade.

The late adolescence study: Charting the development of eating and depressive problems
The primary focus of the Late Adolescence Study was the development of adjustment and psychopathology through young adolescence (mean age = 14.31), mid-adolescence (mean age = 16.03), and the transition from late adolescence into young adulthood (mean age = 22.3). A developmental psychopathological approach was applied to eating and depressive problems in adolescent girls. Girls were initially contacted when they were in either seventh, eighth, or ninth grade. They were enrolled in private schools in a major metropolitan area and were from White, middle to upper-middle class, well-educated families. At each of the three times of assessment, girls reported on their eating attitudes and behaviors, depressive affect, and body image using standard paper-and-pencil questionnaires. In young adulthood, the young women also completed an interview to assess if they had ever met diagnostic criteria for an eating disorder. This study is perhaps the only investigation of eating problems across the entire adolescent decade.

Despite the wealth of resources available to our sample of girls, they are exactly the group most likely to have eating problems (i.e., from White middle to upper-middle class families and attend private, usually all-girl schools). In fact, at each time of assessment, during young adolescence, mid-adolescence, and the transition to adulthood, over a quarter of the adolescent girls we studied could be categorized as having a serious eating problem (see Note 4). Although few girls (7%) had a serious problem at all three times across adolescence, nearly 20% of girls had problems that recurred at two of the three times. This indicates that although the rate of disturbed eating was fairly constant across all three times of assessment (about 25% of girls or young women at each time), it was not always the same girls who experienced problems. The rates and patterns of eating problems in this sample confirm that eating problems pose a serious health problem for adolescent girls, many of whom are still growing.

Depressive problems follow a different pattern than eating problems. Slightly more girls experienced a depressive problem in mid-adolescence than at other times, with about 25% of the girls having a serious problem with depressed feelings at that time. Depressive problems were less common in the young adolescent and young adult periods, with about 17% to 18% of girls having a serious problem. Only 15% of girls reported a serious depressive problem more than once across adolescence, and only one girl was depressed every time she participated in the study

Most girls we studied did not have a serious eating or depressive problem, and their body image was generally positive. In fact, as girls got older their feelings about their bodies became slightly more positive. In young adolescence, 18% of girls reported negative body images in comparison to only 7% reporting negative feelings in mid-adolescence and 5% in young adulthood.

Recurrence of problems
As indicated, only one girl was consistently depressed, and most girls’ eating problems were not continual throughout adolescence. Many girls had difficulties at one time in adolescence but recovered, and their problems did not recur. In addition to concern over who developed serious adjustment problems in adolescence, we were particularly interested in identifying the factors that were associated with recurrence of these problems. We expected that girls with recurrent problems during young and mid-adolescence would be more likely to develop lifelong problems.

Recurrent eating problems.
Examination of girls who had recurrent eating problems in adolescence in comparison to girls who had single episodes of an eating problem or no problem revealed several unique characteristics of girls with recurrent problems (see Note 5). First, girls with recurrent problems went through puberty earlier than their peers. These girls also had higher percentages of body fat than other girls. The findings indicate that eating problems did occur in response to the physical changes of puberty, but only for girls who experienced these changes earlier than their peers. Girls who develop earlier than their peers are gaining weight and growing in height at a time when other girls are still thin, and they are the ones who experience the poorest feelings about their bodies. The fact that these girls actually have higher percentages of body fat than their peers when they are 14 and 16 years of age further supports the suggestion that the unhealthy dieting practices many adopt are in response to the rapid weight changes during puberty.

Second, girls with recurrent eating problems also experience many other adjustment difficulties during adolescence. They report greater externalizing (acting out) and internalizing (moodiness, self-blame, withdrawal) problems than other girls and are more likely to have serious functional impairment resulting from their problems. Interestingly, girls who had eating problems in young adolescence but had recovered by mid-adolescence still suffered some psychological distress, as indicated by increased reports of symptoms such as perfectionism.

Finally, girls with recurrent eating problems in adolescence were more likely than other girls to report an eating problem as young adults — 6 years later! Even as young adults, women who had recurrent eating problems in adolescence continued to have higher percentages of body fat as young adults. They also persisted in having more symptoms of other psychological problems than other young women. In contrast, if girls had single episodes of eating problems fairly early in adolescence (age 14), they were not likely to experience them again as young women. Young women who had positive adjustment throughout adolescence were those who had better body images in adolescence and experienced the physical changes of puberty at the time that was most normative for girls. Consistently healthy adjustment was also associated with positive and warm relationships with parents.

Recurrent depressive problems.
Examination of the recurrence of depressive problems revealed very different behavioral patterns during adolescence. We examined the factors that differentiated girls who had recurrent depressive episodes, single episodes, and positive adjustment. Positive adjustment was again characterized by having more positive family and peer relationships. Whereas the physical changes of puberty resulted in unhealthy eating behaviors for girls who experienced these changes earlier than peers, puberty was not associated with single or recurrent depressive episodes. Instead, single depressive episodes during adolescence were related to situation-specific factors. Girls who experienced these reported negative events having occurred in their lives — most often problems with peers — about the times that they had the depressive problems. Several other studies have also found that depressive episodes in adolescence are often associated with the experience of negative life events, especially those that involve someone close to the adolescent (e.g., having a fight or falling-out with a friend, breaking up with a girlfriend or boyfriend). Whereas in this study single episodes of depression seemed to result from these types of negative experiences, girls with recurrent depressive problems during adolescence did not report experiencing more negative life events than other girls. Instead, these girls reported greater internalizing behaviors separate from the symptoms of their depressive problem.

These internalizing patterns of behavior persisted across the transition from adolescence to young adulthood. Specifically, it was apparent that some girls experienced depressive problems during adolescence and then recovered (or “bounced back”), whereas others demonstrated more continuity in the experience of depressive problems. We looked at the elements of bouncing back by comparing girls who experienced a depressive problem during adolescence and either did or did not have a depressive problem during young adulthood. Again, girls who did not recover had higher reports of other symptoms during adolescence, both more internalizing and externalizing behaviors, even though during adolescence they did not report more depressive symptoms than the girls who bounced back. Interestingly, young women who experienced a depressive problem for the first time as young adults (a single episode) again reported more negative life events as adolescents than young women who always had more positive adjustment.

Girls with both eating and depressive problems
We have described the experiences of having eating problems or depressive problems during adolescence. Clearly, a significant number of girls experienced both types of problems during adolescence, either at the same time or with one problem preceding the other problem. Problems that occurred at the same time are considered comorbid. Examinations of comorbid versus single conditions revealed interesting differences between having a depressive disorder and having an eating disorder. First, eating problems more often preceded a depressive problem rather than the other way around. In some cases, instead of switching symptoms or problems, girls added the depressive problem to an ongoing eating problem. That is, girls with recurrent eating problems were most likely to have depressive problems by mid adolescence.

Second, social relationships were impaired by having a depressive problem to a greater extent than by having an eating problem. Specifically, depressed girls reported poorer relationships with peers and poorer family relationships. Girls with eating problems had significant social impairments only if they also reported high levels of depressive affect. Eating problems alone did not seem to disturb girls’ feelings about their relationships to the extent that feeling depressed disturbed these relationships. However, girls with comorbid problems had the greatest social impairment.

Implications for intervention
What are the implications of these behavioral patterns for practitioners and researchers who are concerned with improving the health and well-being of adolescent girls? Clearly, the biological, social, and psychological changes of adolescence are associated with the development of eating, depressive, and body-image problems for adolescent girls. The specific risk factors differ by problem behavior. Perhaps at puberty a pattern of eating behavior is set in motion with the subsequent trajectory defined by pubertal changes in body image. This implication has direct relevance to the question of when to intervene to offset a pathological trajectory. As early maturing girls are particularly at risk for eating disorders, it is important to begin preventative steps in the middle childhood or elementary school years. Early maturing girls may begin breast development as early as third grade and may reach menarche in fourth or fifth grade. Hence, although their negative responses to their physical development may begin quite early, the development of the eating problem may only occur after these girls face the additional pressures of adolescence, pressures such as school changes and dating.

The reports of eating problems coexisting with depressive affect must also be considered. Girls with both conditions have a preponderance of poor relationships with others, both peers and family members. It follows that these girls will have fewer supportive persons than better adjusted girls to help them cope with their problems. It may be that girls with eating problems who experience poor social relations and hence have fewer social supports are those girls who become depressed. Identifying the existence of the second problem in comorbid situations is particularly important for expanding treatment to address the range of problems that these girls experience.

Girls’ experiences in adolescence are predictive of their experiences as young adults. Problems in young adulthood were often a recurrence or continuation of a behavior pattern that was established in adolescence. This has important implications for those individuals in a position to work with young girls; improving the experiences of young adolescents may result in positive, lifelong consequences.

Author's notes

1. The authors were supported by grants from the National Institute of Child Health and Development (NICHD; Nos. HD24770, HD32376) during the writing of this article. We would also like to acknowledge the support of the W. T. Grant Foundation for support of the Adolescent Study Program from which findings are reported and the NICHD Research Network on Child Well-Being.

2. The authors wish to acknowledge the contributions of Ilana Attie for her work on previous phases of the study; Roberta Paikoff, Richard Fox, Claire Holderness, Alice Michael, and Jamie Traegger, for their assistance with the follow-up project; the staff at Educational Testing Service in Princeton, New Jersey; and the staff of the Adolescent Study Program for their assistance over the course of the project. We are also especially grateful to the young women and their mothers for their continued participation and support in this research.

NOTES

1. The Adolescent Study Program, co-directed by Drs. Brooks-Gunn and Warren with associate director Dr. Graber, has been a collaborative effort of the departments of Gynecology, Pediatrics, and Medicine at the St. Luke’s/Roosevelt Hospital Center in New York, the Center for Children and Families at Teachers College, Columbia University, and the Division of Educational Policy Research at the Educational Testing Service in Princeton, New Jersey. The primary studies conducted under the aegis of the Adolescent Study Program include the Cross-Sectional Adolescent Girls’ Study, the Early Adolescence Study, the Late Adolescence Study, the Mother—Daughter Interactions Study, the Bone Density Study, and the Girls’ Health and Development Project. Several smaller projects have been developed from the primary studies, including the development and implementation of an eating attitudes intervention program (Able To Eat) for early adolescent girls.

2. See Attie and Brooks-Gunn (1995) for a comprehensive review of eating disorders in adolescence.

3. Numbers were drawn from Petersen et al. (1993).

4. The results of the clinical interview that assessed lifetime history of having an eating disorder indicated that 6% of girls met criteria for a frank eating disorder in mid-adolescence and about 5% of the sample met criteria for a disorder during the young adult years. Girls and young women most often were diagnosed with Bulimia; there were fewer cases of Anorexia. The fact that the prevalence of eating disorders was higher in this sample of adolescent girls than the rates for the general population was not surprising given that, based on their economic and environmental backgrounds, these girls are most at risk for having an eating disorder.

5. See Graber, Brooks-Gunn, Paikoff, and Warren (1994) for details on this study.
 

References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.
Attie, I., & Brooks-Gunn, J. (1995). The development of eating regulation across the lifespan. In D. Cicchetti & D. J. Cohen (Eds.), Developmental psychopathology (Vol. 2, pp. 332—368). New York: Wiley.
Graber, J. A., Brooks-Gunn, J., Paikoff, R. L., & Warren, M. P. (1994). Prediction of eating problems: An eight year study of adolescent girls. Developmental Psychology, 30, 823—834.
Petersen, A. C., Compas, B., Brooks-Gunn, J., Stemmler, M., Ey, S., & Grant, K. (1993). Depression in adolescence. American Psychologist, 48(2), 155 – 168.
Simmons, R. G., & Blyth, D. A. (1987). Moving into adolescence: The impact of pubertal change and school context. New York: Aldine.